several grounds, however, a different solution may be preferred, notably providing services free or at less than marginal cost. The following arguments are presented:
1. As the provision of birth control services has many common elements with ordinary medical and public health services and as the latter for various reasons is often socialized, a unified treatment for the supply of all such services may be considered natural and/or preferable.
2. The inefficiency resulting from nonprice (or nominal-price) distribution may be considered negligible. Three points are pertinent here. First, all families are engaged in reproduction; hence it can be expected that the benefits will be broadly spread, affecting at some time or another virtually every family. As a consequence, redistributional effects will generally be moderate. Second, the intrinsic nature of the services is such that demand per family is physically constrained; hence, unlike free transportation or even free aspirin, the allocative inefficiency created by a low price, or by no price, will be small. Third, the cost per person may be low and free supply administratively more advantageous.
3. Any distributional effects that are involved in a subsidized system may be considered positive because such a system would extend to the poor services that were previously available only to the better-off. The same argument could justify the application of discriminatory pricing, i.e., collecting user charges set by the ability to pay, determined by some appropriate yardstick.
Given the socialized distribution of birth control services, the financing of research and development to improve the efficiency, safety, and acceptability of contraceptive techniques must, by necessity, also be collectively provided. Given the assumption of no externalities, the case for each financing is not that it will decrease fertility (although such decrease will be an inevitable by-product, unless compensated for by a reduction of involuntary sterility), just as the case for finding a remedy for headache need not appeal to beneficial effects on labor productivity (although that, too, would be a likely by-product). Rather, insofar as families have any desire to limit their natural